Proteolysis of V antigen from Yersinia pestis

Microbial

Pathogenesis

Proteolysis

1987; 2: 49-62

of V antigen

from

Yersinia

Robert R. Brubaker,” Allen K. Sample, Dong Zheng 2ahorchak.S Ping Chuan Hu§ and Janet M. Fowler Department of Microbiology Ml 488241701. U.S.A. (Received

June 25,1986;

and Public

accepted

Health,

Michigan

pestis Yu,t

Robert

State University.

in revised form August

J.

East Lansing,

11,1986)

Brubaker, R. R. (Dept. of Microbiology, Michigan State University, East Lansing, MI 488241101, U.S.A.), A. K. Sample, D. Z. Yu, R. J. Zahorchak, P. C. Hu and J. M. Fowler. Proteolysis of V antigen from Yersinia pestis. Microbial Pathogenesis 1987, 2: 49-62. Lcr-plasmids of yersiniae are known to mediate a unique low calcium response characterised by restriction of growth in vitro with induction of putative virulence factors including yersiniae outer membrane-peptides (YOPs) and V antigen (Lcr’). A medium was developed that permitted expression of high yields of V by Yersinia pestis KIM in large fermenter vessels. lmmunoblots of specific precipitates prepared by prior molecular sieving showed that native unaggregated V exists as a monomeric 37000 dalton peptide. Fractionation by precipitation with (NH,),SO, and chromatography on phenyl-Sepharose, DEAE cellulose, Sephacryl S200, calcium hydroxyapatite, and Sephadex G200 yielded highly purified antigen as judged by sodium dodecyl sulfate-polyacrylamide gel electrophoresis of parallel preparations from Lcr+ and Lcr- yersiniae. However, yields of V obtained by this process were unexpectedly low. As determined from immunoblots with monospecific polyclonal and monoclonal anti-V, this loss of activity occurred as a function of evident degradation at every step of purification yielding antigenic fragments of about 36 000, 34000, 31 000, 30000, and 28000 daltons. Neutral or acidic pH favored hydrolysis; insignificant cleavage occurred in viable Lcr+ cells of Y. pestis or in culture supernatant fluids. V in neutral cytopolasm from Yersinia pseudotuberculosis or Yersinia enterocolitica did not undergo comparable degradation. Key words; Yersinia; peptides (YOPs).

low calcium

response;

V antigen;

autoproteolysis;

yersiniae

outer membrane

Introduction Wild type cells related Yersinia

calcium response” deficient

medium

the causative agent of bubonic plague, and closely and Yersinia enterocolitica exhibit a unique ‘low

of Yersinia pestis, pseudotuberculosis

characterised (Lcr’).

This

by restriction

phenomenon,

of growth which

at 37°C in enriched

is promoted

by a common

Ca”low

*Address correspondence to: Dr Robert R. Brubaker, Department of Microbiology and Public Health, Giltner Hall, Michigan State University, East Lansing, Ml 48824-l 101, U.S.A. tPresent address: Institute of Epidemiology and Microbiology, China National Center for Preventive Medicine, Beijing, China. SPresent address: Department of Biological Sciences, The University of Alabama in Huntsville, Huntsville, AL, U.S.A. §Present address: Department of Pediatrics, School of Medicine, The University of North Carolina at Chapel Hill, Chapel Hill, NC, U.S.A. 0882-401

O/87/01

0049+14

$03.00/O

0

1987

Academic

Press

Inc.

(London)

Ltd.

50

R. R. Brubaker

et a/.

calcium response or Lcr plasmid of about 45 M Da,.2,3 reflects an ordered metabolic step-down involving decrease of nucleoside triphosphate pools, reduction of adenylate energy charge, and shut-off of stable RNA synthesis.4,5 However, restricted Yersiniae undergo full induction of putative Lcr plasmid-mediated virulence factors6,’ including V and W antigen’ and certain yersiniae outer membrane peptides, termed YOPs.’ Addition of 2.5 I?IM Ca” to restricted organisms or shift of the latter to 26°C promotes growth with repression of V, W, and pertinent YOPS.~,~,~ Mutation to Lcr- eliminates the temperature dependent nutritional requirement for Ca*‘, prevents or significantly reduces expression of V, W, and YOPs, and results in aviruIence.‘“~” This event can reflect loss of the Lcr plasmid’O,” or occurrence of mutations within a regulatory region of this structure.‘.2.‘2,13 YOPs of Y. pseudotuberculosis are distinguished by size as YOPI, YOP2, YOP3, YOP4, and YOP5 with molecular weights of 150000, 44000, 40000, 34000, and 26000 daltons, respectively.g YOPs of Y. enterocolitica are similar except that YOP4 is 36 000 daltonsg These structures are not detected in Lcr+ cells of Y. pestis cultivated under restrictive conditions that promote their synthesis at high titer in Y. pseudotuberculosis and Y. enterocolitica.3.‘4 Nevertheless, antibodies directed against YOPs exist in sera of man and animals during convalescence from plague.‘,15 This observation suggested that significant YOPs are produced by Y. pestis during growth in vivo where they may fulfill an important role in promoting virulence.” YOPl serves as a hydrophobic fibrillar adhesin that accounts for the phenomena of autoagglutination and hemagglutination.’ This structure is not required for expression of virulence and its production, unlike V, W, and the smaller YOPs, may not be fully repressed by Ca”.” Physiological functions of the remaining YOPs are uncertain; plausible roles have been proposed in promoting resistance to complement-dependent’8,‘g and oxygen-dependent*’ mechanisms of killing. V and W antigens are expressed in vitro by all Lcr’ species of Yersiniae” but the specific activity of V in Y. pestis was slightly higher than that from the other Yersiniae.*’ Unlike YOPs, V was not detected in outer or inner membranes but the antigen accumulated in culture supernatants and cytoplasm.‘4,22 W was defined as a 180000 dalton lipoprotein** and V from spent culture medium was characterised by molecular sieving as a 90000 dalton protein.** In contrast, cytoplasmic V was identified by denaturing electrophoresis as a 37000 dalton peptide.14 Anti-W did not provide immunity against experimental plague ** although anti-V conferred passive protection against all Lcr+ species of Yersiniae.22,23 This observation plus the finding that crude V from cytoplasm prolonged survival of Lcr yersiniae in vivo23 suggested that the antigen may serve to directly promote virulence. The purpose of this report is to introduce a method permitting preparation of highly purified V from Y. pestis, although at low specific activity, and to describe its unexpected ability to undergo extensive evident autoproteolysis. Results Requisite for characterising V in Y. pestis was development of a medium both suitable for use in large fermenter vessels and capable of supporting maximum induction of the antigen. Results of preliminary studies, in accord with early observations,‘j showed that a variety of natural products including yeast extract and various protein hydrolysates promoted excellent synthesis provided that Ca*’ was absent, Mg2+ was brought to 20 mM, a source of fermentable carbohydrate was present, pH was maintained above neutrality, and the culture received adequate aeration. These findings permitted formation of a medium that provided V in crude cell extracts at a specific

V antigen

of Yersinia

Fig. 1. Growth were shifted from

51

of Y. pestis KIM (0, Lcr’, 0. Lcr-) in oxalated 26°C to 37°C at an optical density of 1 .O.

Sheffield

NZ amine.

Fermenter

vessels

activity (-1 unit per mg protein) equivalent to that obtained with the complex chemically defined medium utilised previously. 24 The specific activity of extracellular V was similar although spent medium usually required concentration before the antigen became detectable via precipitation in agar. Growth of Lcr+ and Lcr- cells of Y. pestis in fermenter vessels containing 10 I of this medium is shown in Fig. 1; bacteria were harvested for preparation of V after restriction became complete. Initial observations showed that separation of partially purified V into precise fractions was favored at slightly alkaline pH in the presence of free sulfhydryl groups. Attempts to demonstrate an effect of common metallic cations or metal chelators on storage of the antigen were not successful. Accordingly, all buffers initially used for purification were prepared between pH 7.7 and 8.0 and contained 1 .O mM dithiothreitol. The response of V in these buffers during fractionation of cytoplasm on various chromatographic columns was determined in order to devise an appropriate method of purification (Fig. 2). Sizing by molecular sieving yielded a broad plateau of precipitin activity in samples ranging from about 100 000 to 60 000 daltons followed by a peak averaging 37 000 daltons (Fig. 2(a)) Many other putative virulence factors within cytoplasm including W antigen, murine exotoxin,25 and antigen 426 are in excess of 100 000 daltons and thus were separated without overlap from V by this process. To determine the size of native V, samples of recovered antigen ranging from 100 000 to 37 000 daltons were reacted with excess anti-V and the washed precipitates were subjected to sodium dodecyl sulfate-polyacrylamide gel, electrophoresis (SDS-PAGE). Other than heavy and light chains of y-globulin, the only peptide detected in stained gels was a 37 000 dalton structure (Fig. 3) that also reacted with anti-V in immunoblots (not illustrated). Similar molecular sieving with buffer containing 0.15 M NaCl favored elution of V as an apparent 90 000 dalton component with decreased appearance of the 37 000 dalton structure. This shift was enhanced by use of buffer with 0.5 M NaCl where the antigen was almost entirely eluted as a fraction of 90000 daltons (not illustrated). Conditions used for chromatography on phenyl-Sepharose did not permit absorption of V (Fig. 2(b)) although the exotoxin and much bulk protein was removed by this

52

R. R. Brubaker

et al

Fig. 2. Purification of V antigen on (a) Sephadex G200, (b) phenyl-Sepharose CL-4B, (c) DEAE cellulose, and (d) calcium hydroxyapatite. Eluted macromolecules were monitored by absorbance at 280 nm (0) and V antigen was determined by diffusion in agar against rabbit polyclonal monospecific antiserum (m). Buffers used were 0.05 M Tris. HCI, pH 7.8 (a); 0.05 M Tris. HCI, pH 7.8 with change to distilled water at arrow (b); 0.05 M Tris. HCI, pH 7.8 with introduction at arrow of gradient (dashed line) in same buffer of NaCl (zero to 0.5 M) (c); and 0.01 M sodium phosphate, pH 7.7 with introduction at arrow of gradient (dashed line) of same buffer 0.01 M to 0.5 M) (d). Molecular weight markers in (a) are given in order of decreasing size: bovine thyroglobulrn (670000 daltons), bovine gamma globuljn (158000 daltons). chrcken ovalbumin (44 000 daltons), horse myoglobin (17 000 daltons), and vitamin B-1 2 (1350 daltons) (A).

process. The antigen was absorbed on DEAE cellulose and subsequently released at about 0.1 M after initiation of a linear concentration gradient ranging from 0 to 0.5 M NaCl (Fig. 2(c)). This procedure separated a neutral brown pigment prior to application of the gradient; W antigen and exotoxin were eluted at about 0.2 M NaCl in samples that partially overlapped V and immediately preceded appearance of nucleic acids. Exotoxin and other large proteins dissociated into subunits approximating V in molecular weight during chromatography on DEAE cellulose. V in cytoplasm dialysed against sodium phosphate buffer at low ionic strength (0.01 M) was absorbed on calcium hydroxyapatite and then released at about 0.15 M after initiation of a linear gradient of the same buffer (0.01 to 0.5 M) (Fig. 2(d)). The antigen was effectively concentrated by this process although significant overlap with exotoxin occurred. The brown pigment and W antigen were irreversibly absorbed during chromatography on calcium hydroxyapatite.

V antigen

of Yersinia

53

-66.2

- 21.5

- 14.4 I

2

M

Fig. 3. Coomassie blue-stained 12.5% SDS-PAGE of sized 100000 (lane 1) and 37 000 (lane 2) dalton immunoprecipitated samples from Fig. 2(a); hc and Ic are heavy and light chains of y-globulin, respectively. Marker peptides (lane M) were (top to bottom) phosphorylase (92500 daltons), bovine serum albumin (66200 daltons), ovalbumin (45000 daltons), carbonic anhydrase (31 000 daltons), soybean trypsin inhibitor (21 500 daltons), and lysozyme (14400 daltons).

The amount of V recovered after chromatography by any one of these four methods was about 90% of that initially present. However, significant losses occurred during purification by any combination of two or three of these procedures, regardless of order, permitting recovery of less than 10% of the starting activity. This phenomenon is shown in Table 1 where V was initially concentrated by precipitation with (NH4)$S04 before chromatography on phenyl-Sepharose, DEAE cellulose, Sephacryl S200, calcium hydroxyapatite, and then Sephadex G200. An identical process of fractionation was undertaken with an extract of isogenic Lcr- organisms in order to provide analogous samples. Comparison of the two series via SDS-PAGE showed that the final Sephadex G200 sample obtained from Lcr~ cells contained 12 major peptides (Fig. 4(a)). These structures plus two additional peptides were present in the comparable preparations obtained from Lcrf organisms (Fig. 4(b)). The largest peptide unique to the latter was 37 000 daltons and thus assumed to represent V. lmmunoblots were performed on these preparations with rabbit polyclonal anti-V in order to verify the location of V and to resolve the nature of the smaller Lcr’-specific structures. Other than the common 66000 dalton peroxidase activity, the only component detected in whole Lcrf organisms was a 37000 dalton peptide equated with V (Fig. 5(a)). However, during disruption of the bacteria and therafter, additional peptides of about 36 000, 34 000,31 000, 30 000, and 28 000 daltons were generated

54

R. R. Brubaker

Table 1

Purification KIM and corresponding

of V antigen from crude ceil-free extracts samples from an isogenic Lcrrmutant

Vol. Lcr

Preparation

Total protein (w)

Protein

(ml)

(w/ml)

+

Crude extract 20 to 50% (NH&SO, Phenyl-Sepharose DEAE cellulose Sephacryl S200 Ca hydroxyapatite Sephadex G200

60 19 115 55 15 17 95

31 .l 50 6.5 3.9 2.3 0.94 0.11

1866 950 747.5 215 34.5 16 10.5

-

Crude extract 20 to 50% (NH,),S04 Phenyl-Sepharose DEAE cellulose Sephacryl S200 Ca hydroxyapatite Sephadex G200

33 24 117 51 16 58 96

38.8 40.6 8.0 5.2 4.8 0.6 0.1

1280 974 936 265 76 35 9

(U pir

ml)

80 200 30 40 70 40 4

of Lcr+ Yersinia

et al

pestis

%

Total v (U)

Specific activrty

Recovery

4800 3800 3450 2200 1050 680 380

2.6 4.0 4.6 10.3 30.4 42.5 36.4

100 79 72 46 22 14 8

that reacted with anti-V. Some of these structures may represent the smaller Lcr’specific peptides detected by staining (Fig. 4(b)). These components were absent in comparable samples obtained from Lcr- Yersiniae (Fig. 5(b)). Since the appearance of these peptides was dependent upon the presence of V, the latter was assumed to serve as a common precursor. To further define this relationship, a mouse monoclonal

(0)

(b)

I

2

3

4567M

I

2

3

4

5

67

Fig. 4. Coomassie blue-stained 12.5% SDS-PAGE preparation of sequential antigen from Lcr(a) and Lcr’ (b) Y. pestis KIM. Whole cells (lane l), crude phenyl-Sepharose CL-4B (lane 3). DEAE cellulose (lane 4), Sephacryl S200 (lane (lane 6), and Sephadex G200 (lane 7). Marker peptides (lane M) are as in Fig. 3. (37 000 daltons) and a putative degradation fragment (31 000 daltons).

steps of punfrcation of V cell-free extract (lane 2), 5), calcium hydroxyapatite Arrows indicate V antigen

V antigen

of Yersinia

55 (a)

(b)

92.5 -

21.5-

1234567

I2

345

6

7

Fig. 5. lmmunoblots prepared with rabbit polyclonal anti-V of the same preparations of Lcr’ (a) and Lcrr (b) Y. pestis KIM shown in Fig. 4. The 66000 dalton peptide is a peroxidase activity common to both cell types and served as an internal marker. Remaining peptides are Lcr’-specific and consist of V (37000 daltons) and evident degradation products of about 35 000, 33 000, 31 000, 28 000, and 27 000 daltons.

antibody to V was used to prepare immunoblots of the same samples previously reacted with the polyclonal antiserum. This antibody typically exhibited minor interractions with elements of both Lcr+ and Lcr- preparations and strongly reacted with an epitope shared by V and the 31 000 dalton peptide (Fig. 6). This result would be anticipated if (a)

(b)

1234567

Fig. 6. lmmunoblots prepared with mouse monoclonal Lcr- (b) Y. pestis shown in Fig. 4. The 31 500 peptide recognised on V antigen. Remaining detectable peptides dalton degradation fragment.

1234567

anti-V of the same preparations of Lcr’ (a) and detected in both cell types shares the epitope include V per se (37 000 daltons) and its 31 000

56

R. R. Brubaker

et al

the epitope in question was located near one end of the antigen where it would be lost upon cleavage of the 1000 and 3000 dalton fragments that cause formation of the other degradation products. Results of further study showed that V did not undergo significant degradation in cell-free extracts maintained at pH 9.0 whereas cytoplasm buffered at pH 8.0 or less exhibited hydroylsis typical of that shown in Fig. 5. The presence of dithiothreitol was not required for degradation although the percent recovery of V upon purification via the schema shown in Table 1 was doubled by use of buffers lacking the reductant. The antigen present in supernatant fluid of stationary phase cultures existed primarily as the 37000 dalton peptide although the 31 000 dalton structure was also detected as a minor component (Fig. 7). The polyclonal antiserum detected an additional approximate 45 000 dalton Lcr’-specific peptide in culture supernatant fluids that was not observed in cytoplasm. To determine if V from other Yersiniae could also undergo hydrolysis, we prepared immunoblots of whole cells and cytoplasm of additional strains of Y. pestis and of restricted isolates of Y. pseudotuberculosis and Y. enterocolitica. The antigen from all of three tested Lcr’ isolates of Y. pestis exhibited degradation during storage at neutral pH whereas that from eight Lcr+ strains of Y. pseudotuberculosis and Y. enterocolitica remained intact. This phenomenon is illustrated in Fig. 8 with two selected strains of each of the three species.

Discussion

and conclusions

Important problems to be resolved concerning the role of Lcr plasmids in promoting disease by yersiniae include determining the physological roles of V and the smaller YOPs and defining why the latter are not expressed by cells of Y. pestis cultivated under restrictive conditions. As previously noted, evidence favoring a role of V in mediating virulence is indirect and based primarily on the observation that the antigen promotes survival of Lcrr mutants in vivo and that anti-V confers passive immunity to

(b) -

92.5

-

66.5

21.5

ml23456m

I23456

123456

Fig. 7. Coomassie blue-stained 12.5% SDS-PAGE (a), rmmunoblot prepared with polyclonal rabbit (b) or monoclonal mouse (c) antiserum of whole cells (lane l), crude extract (lane 2), and supernatant (lane 3) of Lcr’ and supernatant (lane 4), crude extract (lane 5) and whole cells (lane 6) of Lcr- Y. pestis KIM. Marker peptides (lane M) are as in Fig. 3.

V antigen

of Yersinia

57

66.2

-

45.0-

31.0 -

31.5-

Fig. 8. lmmunoblot added to SDS-PAGE at room temperature

prepared with rabbit polyclonal anti-V of Lcr’ whole cell extracts. Extracts were sample buffer immediately after sonication (lanes a, c, e, g, i and k) or after 36 hours (lanes b, d, f, h, j and I). Strains tested are Y. pseudotuberculosis PBI (lanes a and b), Y. pseudotuberculosis YPIII (lanes c and d), Y. enterocolitica WA (lanes e and f), Y. enterocolitica Ml (lanes g and h), Y. pestis KIM (lanes i and j), and Y. pestis EV76 (lanes k and I).

Lcr+ organisms. 23 Although V was discovered in 1956,8 little is known about the structure and function of the antigen due to its known instability.** Our attempts to devise improved methods to concentrate V have generally proven to be ineffective thus we initiated this study to determine the cause of the evident lability. The resulting observation that the antigen undergoes extensive hydrolysis provides an explanation for its instability. The existence of this proteolytic activity in Y. pestis and its evident absence in Y. pseudotuberculosis and Y. entercolitica may also account for the absence of detectable YOPs in the former. Curiously, proteolysis of V was not detected in whole cells nor did extensive degradation occur in culture supernatant fluids. An explanation for this stability will probably require identification of the salient proteolytic activity. Should this function be catalysed by one or more enzymes distinct from V, then stability in whole cells and in spent medium could reflect compartmentalisation. Alternatively, if the proteolytic activity is an intrinsic property of V per se, then stability in whole cells might involve the presence of inhibitors or maintenance of the antigen in some non-hydrolytic configuration. Since V was stable in alkaline buffers, lack of proteolysis may merely reflect the mild alkaline pH of about 8.5 observed for spent culture medium. Evidence favoring an intrinsic proteolytic capability of V consists of the observation that similar degradation products were generated throughout the process of purification. This observation is consistent with hydrolysis by V per se since it seems unlikely that a distinct proteolytic activity could undergo co-purification by this procedure.

R. R. Brubaker

58

et al.

However, the final product was clearly not homogenous thus the alternative remains that a separate protease accounts for degradation. This possibility is strengthened by the observations that extracellular V from Y. pestis and intracellular V from Y. pseudotuberculosis and Y. enterocolitica did not undergo comparable degradation. Additional study will be required to resolve this relationship. In any event, it is evident that V is a potentially labile molecule and that appropriate methods of stabilisation may have to be developed in order to define its function. However, this study has provided useful information regarding the structure of V. By use of molecular sieving, we obtained a value of 37000 daltons for the native antigen as opposed to the previously reported molecular weight of 90000 daltons.” The former is identical to that observed by electrophoresis in denaturing gels15 and thus represents a monomeric molecular weight. It was therefore of interest to determine if V in samples containing larger sized molecules was composed of additional subunits and, if so, were they identical or distinct. As judged by results obtained by SDS-PAGE of specific immune precipitates, the antigen in these samples was entirely composed of the 37000 dalton peptide and possibly its degradation products. Furthermore, the proportion of apparent higher molecular weight V could be increased by elevating the concentration of NaCl in the eluting buffer. These observations are inconsistent with the existence of the putative 90 000 dalton structure as a dimer or trimer of the 37 000 dalton peptide or as a combination of the latter with a distinct peptide. However, the evident larger antigen could reflect a loose aggregate of 37000 dalton peptides and their degradation products promoted by increased ionic strength. A number of proteins, including the bacteriocin pesticin,27 have been purified to near homogeneity from cytoplasm of Y. pestis without evidence of proteolytic destruction. Therefore, the discovery that V undergoes degradation was not anticipated. This realisation presents a number of new variables that may influence the nature of the low calcium response of yersiniae. For example, many important exotoxins are known to undergo hydrolytic cleavage during a process of activation.** The possibility now exists that V or other virulence functions of Y. pestis may also require proteolytic activation to exhibit biological activity. Further studies designed to resolve these possibilities are presently in progress.

Materials

and methods

Bacteria. Unless stated otherwise, materials prepared from Lcr’ or Lcr- cells of Y. pestis strain KIM were used in all experiments. These isolates are non-pigmented” and thus are only conditionally virulent (via intravenous injection in normal mice”“). Salient properties of these and other isolates used for determining degradation of V are shown in Table 2. Cultivation. Yersiniae were cultivated for preparation of V in a modification of the previously described oxalated medium.‘j To prepare 10 I of this fermenter medium, 300 g of Sheffield NZ amine, Type A (Kraft, Inc., Memphis, TN) were dissolved in 2000 ml of 0.5 N NaOH containing 0.04 moles of sodium oxalate. After storage overnight at 5°C. precipitated calcium oxalate was removed by filtration or low speed centrifugation and the solution was brought to 9600 ml in a fermenter vessel by addition of distilled water. After sterilisation by autoclaving and cooling, the medium received 200 ml of sterile 2.0 M MgCI, and 200 ml of sterile 1.0 M potassium gluconate. Organisms previously grown for 2 days at 26°C on slopes of Tryptose blood agar base (Difco Laboratories, Detroit, Ml) were removed in 0.033 M potassium phosphate buffer, pH 7.0 and used to inoculate 100 ml of fermenter medium contained in a 1000 ml Erlenmeyer flask. This first transfer was aerated for 12 h at 26°C on a model R26 reciprocating floor shaker (New Brunswick Scientific Co. Inc., New Brunswick, NJ) at 200 rpm and then used to inoculate second transfers of the same medium (200 ml per 2000 ml flask). After similar incubation, these cultures were used to directly inoculate medium within fermenter vessels at an optical density

V antigen of Yersinia

59

Table 2 Yersinia used in determination antigen in crude extracts at neutral pH Species Y. pestis

Y. pseudotuberculosis

Y. enterocolitica

Strain KIM EV76 M23 PBI EP2 MD31 Nelson Hale Galligue Parkin YPIII WA E661 E701 E736 Ml M3 M4 M5

Biotype or serotype Mediaevalis Orientalis Orientalis I II III IV I I I III 0:8 0:8 0:4,32 0:21 0:3 0:3 0:3

0:15

of stability

of V

Source M. J. Surgalla M. J. Surgalla T. W. Burrows T. W. Burrows T. W. Burrows J. Marshall T. W. Burrows T. W. Burrows T. W. Burrows T. W. Burrows H. Wolf-Watz P. B. Carter D. A. Schiemann D.A. Schiemann D. A. Schiemann A. Mellado A. Mellado A. Mellado A. Mellado

(620 nm) of about 0.4. The latter were inserted into a Model MF-214 fermenter (New Brunswick Scientific Co. Inc.) and the contained medium received mechanical stirring and aeration at 500 rpm and 12 I per min, respectively. Growth was initiated at 26°C and then the temperature was adjusted to 37°C when the optical density was about 1. Crude antigen. Fermenter vessels containing restricted Lcr+ yersiniae or Lcr- mutants in late log phase were stored in the cold while bacteria were removed from portions of the medium by centrifugation at 11 000 g for 30 min at 4°C. The culture supernatant fluid was used as a source of extracellular V following sterilisation either by exposure to CHCI, vapor or by passage through a bacteriological membrane filter, precipitation by saturation with (NH4).S04, centrifugation at 27 000 g for 20 min. and dialysis of the resulting sediment against 0.01 M Tris * HCI buffer, pH8 containing 1 .O tT?M dithiothreitol. The pelleted bacteria were suspended in 0.033 M potassium phosphate buffer, pH 7.0, centrifuged as previously described, and then suspended in 0.05 M Tris. HCI, pH 7.8. This preparation was treated for four intervals of 15 set with an ultrasonic probe (MSE Ltd., London, England) and cellular debris was removed by centrifugation (12000 g for 15 min); the resulting clarified supernatant was sterilized by exposure to CHCI, vapor. lmmunoprecipitates. Following separation of crude antigen on Sephadex G200, aliquots of 0.25 units of V were removed from individual fractions and mixed with 0.40 units of polyclonal monospecific anti-V. Precipitins were allowed to form at 37°C for 1 hour followed by 18 hours at 0°C in an ice-water bath. lmmunoprecipitates were collected by centrifugation at 12000 g for 30 min and washed three times in 0.05 M Trisv HCI, pH 6.8 containing 150 mM NaCI. The immunoprecipitates were resuspended in SDS-PAGE sample buffer of Laemmli33 and were separated on 12.5% SDS-polyacrylamide gels. Purification of V antigen. Unless stated otherwise, all buffers used for dialysis or chromatography contained 1 .O mM dithiothreitol and separations were performed at 4°C. Samples no larger than 100 ml were applied to the surface of a column 2.5 x 15 cm) containing phenylSepharose CL-4B equilibrated in 0.05 M Tris* HCI buffer, pH 7.8. Unabsorbed material was eluted with the same buffer and then distilled water followed by 8 M urea was used to elute weakly and strongly absorbed hydrophobic material, respectively. Molecular sieving with Sephacryl S200 (Pharmacia Inc., Piscataway, NY) was performed with samples not exceeding

R. R. Brubaker

60

et al

5 ml in a column of 2.5 x 100 cm using 0.05 M Tris. l-ICI, pH 7.8 as eluant at a flow rate of 0.75 ml per min. Similarly, Sephadex G200 (Pharmacia, Inc.) equilibrated with the same buffer in a column of 2.5 x 100 cm was used to determine the molecular weight of V at a flow rate of 0.2 ml per min. Molecular weight markers used were thyroglobin, gamma globulin, ovalbumin, myoglobin, and cobalamin of 670 000, 158 000, 44 000, 17 000, and 1350 daltons, respectively. Chromatography on DEAE cellulose (Whatman Inc., Clifton, NJ) equilibrated in 0.05 M Tris . HCI buffer, pH 7.8 was performed in a column of 2.5 x 60 cm at a flow rate of 1 ml per min. After unabsorbed and poorly absorbed material was eluted, a linear gradient of 0 to 0.5 M NaCl in the same buffer was initiated. A column (2.5 x 60 cm) was used for chromatography on calcium hydroxyapatite (Bio-Rad Inc., Richmond, CA) where samples were applied in 0.01 M sodium phosphate buffer, pH 7.7 and, after unabsorbed material was eluted with the same buffer, a gradient to 0.5 M sodium phosphate, pH 7.7 was initiated. Chromatography on calcium hydroxyapatite was performed at room temperature. The use of Sephadex G200 as a final step in purification was not required because it resulted in a reduced recovery and lower specific activity. This final sizing step was used to remove degradation products generated during previous purification steps and to demonstrate that a proteolytic activity was present in the final product. Determinations. Protein was determined by deficient in nucleic acids, by spectrophotometry determined as precipitin activity in linear dilutions by others.22

the method at 260 and of samples

of Lowry et al.3’ or, in samples 280 nm. V and other antigens were by diffusion in agar as performed

Antisera. Polyclonal monospecific anti-V was prepared as described previously23 using material prepared essentially as described here as antigen. This process involved absorption of the antiserum with acetone-dried or lyophilised disrupted whole Lcryersiniae to remove antibodies directed against common determinants and then absorption with live restricted Lcr’ cells of Y. pseudotuberculosis PBl/+ and Y. enterocolitica WA to eliminate possible contamination with anti-YOPs. Remaining y-globulin (IgG) was then purified by fractionation with 50% saturated (NH,),SO,, chromatography on DEAE cellulose, and then chromatography with Sephadex G200. Reagent prepared by this process possessed monospecific anti-V activity23 although some antibody directed against an approximate 45000 dalton extracellular Lcr‘specific peptide was sometimes detected (Fig. 7). Monoclonal anti-V was prepared as described by Hu et a13* using an antigen prepared essentially as described in this report. Antibody (IgG) resulting from the selected hybridoma was purified as described for the polyclonal reagent. Rabbit polyclonal antisera capable of reacting against W, murine toxin, or antigen 4 were obtained from Dr William D. Lawton.

lmmunoblots. Electrophoretic transfer of protems separated on 12.5% polyacrylamide gels according to the method of Laemmli33 was performed essentially as described by Towbin et al.34 The buffer used for subsequent steps contained 50 mM Tris, pH 7.5, 0.05% Tween 20, and 0.87% NaCl (TBS .Tw20). After a preliminary wash, unbound sites on the immunoblot were blocked in 5% fetal calf serum for 2 h. Primary antibody at 3-5 pg/ml was added directly and the immunoblot incubated for an additional 2 h. After five washes in TBS .Tw20, horseradish peroxidase conjugated anti-rabbit or anti-mouse IgG (Sigma, St Louis, MO) was added at a 1 : 2500 dilution, incubation continued for 2 h, and a further five washes in TBS .Tw20 were performed. Prior to development, the immunoblot was washed two times in 50 I?IM Tris. HCI, pH 7.5 containing 10 mM EDTA and 0.87% NaCI. Development was made in this final buffer containing 0.005% H202 and 0.03% 3,3’-diaminobenzidine, and was stopped by washing in distilled water. Use of peroxidase linked secondary antibody resulted in the appearance of a 66 000 dalton band on the immunoblots which did not appear with use of alkaline phosphatase conjugated antibody. Portions of this research provided the basis of the thesis fulfillment of the requirements for the M.S. degree. The technical is gratefully acknowledged. This effort was supported by Public from the National Institute ‘of Allergy and Infectious Diseases. from the Michigan Agricultural Experiment Station.

presented by R.J.Z. for partial assistance of Mary J. Phillipich Health Service Grant Al 19353 This is journal article no. 12021

V antigen

of Yersinia

61

References 1. 2. 3.

Goguen JD, Yother J, Straley SC. Genetic analysis of the low calcium response in Yessinia pestis MIJ dl (Ap /ac) insertion mutants, J Bacterial 1984; 160: 842-8. Portnoy DA, Blank HF, Kingsbury DT, Falkow S. Genetic analysis of essential plasmid determinants of pathogenicity in Yersiniapestis. J Infect Dis 1982; 148: 297-304. Portnoy DA, Wolf-Watz H, Bolin I, Beeder AB, Falkow S. Characterization of common virulence plasmids in Yersinia species and their role in the expression of outer membrane proteins. Infect lmmun 1984;

4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.

43:

108-l

4.

Zahorchak RJ, Charnetzky WT, Little RV, Brubaker RR. Consequences of Ca” deficiency on macromolecular synthesis and adenylate energy charge in Yersinia pestis. J Bacterial 1979; 139: 792-9. Charnetzky WT, Brubaker RR. RNA synthesis in Yersinia pestis during growth restriction in calciumdeficient medium. J Bacterial 1982; 149: 1089-95. Brubaker RR, Surgalla MJ. The effect of Ca” and Mg” on lysis, growth, and production of virulence antigens by Pasteurella pestis. J Infect Dis 1964; 114: 13-25. Portnoy DA, Moseley SL, Falkow S. Characterization of plasmids and plasmid-associated determinants of Yersinia enterocolitica pathogenesis. Infect lmmun 1981; 31: 77582. Burrows TW, Bacon GA. The basis of virulence in Pasteurella pestis: an antigen determining virulence, Br J Exp Pathol 1956; 37: 481-93. Bolin I, Porn-toy DA, Wolf-Watz H. Expression of the temperature-inducible outer membrane proteins of Yersiniae. Infect lmmun 1985; 48: 23440. Brubaker RR. The Vwa-’ virulence factor of Yersiniae: molecular basis of the attendant nutritional requirement for Ca”. Rev Infect Dis 1983; 5: S748-58. Portnoy DA, Martinez RJ. Role of a plasmid in the pathogenicity of Yersinia species. Curr Top Microbial lmmunol 1985; 118: 29-51. Yother J, Goguen JD. Isolation and characterisation of Ca*’ -blind mutants of Yersiniapestis. J Bacterial 1985; 164: 704-I I. Straley SC, Bowmer WS. Virulence genes regulated at the transcriptional level by Ca*’ in Yersinia pestis include structural genes for outer membrane proteins. Infect lmmun 1986; 51: 445-54. Straley SC, Brubaker RR. Cytoplasmic and membrane proteins of Yersiniae cultivated under conditions simulating mammalian intracellular environment. Proc Natl Acad Sci USA 1981; 78: 12248 Mazza G, Karu AE, Kingsbury DT. Immune response to plasmidand chromosome-encoded Yersinia antigens. Infect lmmun 1985; 48: 676-85. Kapperud G, Namork E, Skarpeid H-J. Temperature-inducible surface fibrillae associated with the virulence plasmid of Yersinia enterocolitica and Yersinia pseudotuberculosis. Infect lmmun 1985; 47: 561-6.

17.

Bolin

H. Molecular cloning of the temperature-inducible outer membrane protein 1 of Infect lmmun 1984; 43: 72-8. Pai CH, DeStephano L. Serum resistance associated with virulence in Yersinia enterocolitica. Infect lmmun 1982; 35: 605-l 1. Martinez RJ. Plasmid-mediated and temperature-regulated surface properties of Yersinia enterocolitica. Infect lmmun 1983; 41: 921-30. Lian CJ, Pai CH. Inhibition of human neutrophil chemiluminescence by plasmid-mediated outer membrane proteins of Yersinia enterocolitica. Infect lmmun 1985; 49: 145-51. Perry RD, Brubaker RR. Vwa’ phenotype of Yersinia enterocolitica. Infect lmmun 1983; 40: 166-71, Lawton WD, Erdman RL, Surgalla MJ. Biosynthesis and purification of V and W antigen in Pasteurella pestis. J lmmunol 1963; 91: 179-84. Une T, Brubaker RR. Roles of V antigen in promoting virulence and immunity in Yersiniae. J lmmunol 1984; 133: 2226-30. Zahorchak RJ, Brubaker RR. Elimination of Ca”-dependence in Yersinia pestis by exogenous nucleotides. Infect lmmun 1982; 38: 953-9. Brown SD, Montie TC. Beta-adrenergic blocking activity of Yersinia pestis murine toxin. infect lmmun 1977; 18: 85-93. Crumpton MJ, Davies DAL. An antigenic analysis of Pasteurella pestis by diffusion of antigens and antibodies in agar. Proc Roy Sot 1956; B145: 109-34. Hu PC, Brubaker RR. Characterisation of pesticin: separation of antibacterial activities. J Biol Chem 1974; 249: 4749-53. van Heyningen WE. Bacterial exotoxins. In: Braude Al, Davis CE, Fierer J, eds. Medical Microbiology and Infectious Disease. Philadelphia: W. B. Saunders Co., 1981: 5163. Jackson S, Burrows TW. The pigmentation of Pasteurella pestis on a defined medium containing hemin. Br J Exp Pathol 1956; 37: 577-83. Une T, Brubaker RR. ln vivo comparison of avirulent Vwa- and Pgm- or Pst’ phenotypes of Yersiniae. Infect lmmun 1984; 43: 895-900. Lowry OH, Rosebrough NJ, Farr AL, Randall RJ. Protein measurement with the Folin phenol reagent. J Biol Chem 1951; 193: 265-75. Hu PC, Cole RM, Huang YS et a/. Mycoplasma pneumoniae infection: role of a surface protein in the attachment organelle. Science 1982; 216: 313-5. Yersinia

18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32.

I. Wolf-Watz

pseudotuberculosis

62 33. 34.

Ft. R. Brubaker

et a/.

Laemmli UK. Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 1970; 227: 680-5. Towbin H, Staehelin L, Gordon J. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 1979; 76: 4350-4.